A sensor code that appears on the first drive after a downpipe is fitted is usually read as a sensor failure, and usually it is not one. The sensor was working before the work was done. What changed is the environment it reads: the pipe it sits in, the position of its boss, the gas reaching it, and the temperature of the metal around it.
So establish which kind of code it is, not what the dashboard says about it. A heater circuit code, a slow response code and a range code point at three different things, and treating them as one fault is how people end up fitting sensors repeatedly without solving anything.
The sensor has to sit in the gas stream, reach working temperature, and stay there without being cooked. Its tip produces a small voltage once hot and exposed to exhaust gas, and it compares that against outside air drawn through its body. Anything that changes either side of that comparison changes the reading.
It also needs a clean reference. If the sensor's vent is blocked, greasy, or painted over during the job, the reading drifts and the code that appears looks like a fuelling problem rather than a sensor one.
The first is position. A new downpipe can put the boss at a different point along the pipe, at a different angle, or at a different depth. A shallow boss leaves the tip at the edge of the stream, where it reads slowly and cools between cycles. A deep one puts the tip in the middle of the flow and can overheat it.
The second is leakage. A joint not fully sealed upstream of the sensor draws air in on the pulses and the sensor reads lean. The ECU adds fuel, and the code that appears may be a mixture code rather than anything mentioning the exhaust. A v-band not seated, a gasket left out, or a dry slip joint is enough to do it.
The third is heat. A titanium pipe and a stainless one do not move heat into the boss in the same way, and a thin wall runs hotter than a thick one. That changes how quickly the sensor reaches temperature and how hot it gets, which affects response codes and heater codes alike. Wrapping the pipe changes it again.
There is also what the pipe now contains. If the downpipe removed a converter, or moved the sensor relative to one, the gas it sees is a different composition at a different temperature, and a code calibrated around the original arrangement may no longer make sense. The pipe material is part of that decision.
Start cold. Run a hand along the joints upstream of the sensor while the engine idles from cold, feeling for pulses, and look for soot tracks at the flanges, which leave a dark edge where gas has escaped. A leak that only appears hot will not show this way, so repeat after a drive.
For position, take the sensor out and look at its tip. An even deposit across the shield suggests it sits in the stream; a clean side and a sooted side suggest it sits at the edge. Compare boss depth with tip length with the sensor in your hand, and check that nothing inside the pipe shadows the tip.
If the code is a heater or circuit code, the wiring is the first suspect, because the job moved the pipe and disturbed the loom. Check that the connector is fully seated, that the cable is not stretched tight, and that it has not been trapped against the pipe. A taut cable fails at the plug before the sensor does.
If the code is a response or range code and there is no leak, the fix is usually the boss rather than the sensor: move it, change its depth, or change its angle so the tip sits where the gas is. Replacing a sensor in the same boss buys time instead of solving anything. Boss position belongs in the design of the downpipe, not in the aftermarket.
